Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ames!lll-winken!lll-lcc!rzh
From: rzh@lll-lcc.UUCP (Roger Hanscom)
Newsgroups: comp.sys.ibm.pc
Subject: Logitech mouse on the AT
Keywords: Logitech, mouse, Fortran
Message-ID: <2364@lll-lcc.UUCP>
Date: 16 Mar 89 23:08:32 GMT
Organization: Lawrence Livermore Labs, LCC, Livermore Ca
Lines: 27


  Has anybody out there worked with the Logitech high resolution
  mouse??  It is a bus-type device.  The first attempt at coding
  a mouse routine resulted in the cursor jumping from place to
  place on the screen.  The code reads the current mouse hot-spot
  and moves the cursor there.  Logitech says that this is not the
  way to do it, and sold us their 'mouse toolkit'.  Unfortunately,
  the mouse toolkit is C-callable (MS C), and the application is
  Fortran (don't flame me, I'd write it in C if it were my decision!).
  The mouse toolkit uses an event handler to store mouse x-y's in
  a stack, and then the application can "move" the cursor through
  these x-y's, giving the appearance of smooth motion.  Converting
  the C to Fortran and calling their assembler toolkit event handler
  (function #12) locks the AT up.  Reboot time!  I think that the
  problem is the result of fundamental differences in the way the
  two languages handle addressing and segments (the application
  must pass the assembler stuff the address of a subroutine in the
  application environment and of course, the code needs a place to
  store the mouse x-y's).  We are able to call some of the other
  (less complicated) assembler functions from Fortran with little
  difficulty (just by changing the way the call vector gets unpacked).

  Can anyone help?  ($&@*$+! mouse....where's my hammer?)

=================================================================
    roger       rzh@lll-lcc.llnl.gov
                {ucbvax,ames,...}!lll-lcc!rzh